Abstract LB182: Mutagenesis screens support potential best-in-class profile for selective, brain-penetrant, and TRK-sparing ROS1 inhibitor NVL-520

Abstract

Abstract Introduction. ROS1 fusions are recurrent oncogenic drivers in non-small cell lung cancer (NSCLC). Tyrosine kinase inhibitors (TKIs) crizotinib, entrectinib, and repotrectinib have been approved for the treatment of ROS1-positive NSCLC. Response to these TKIs can be limited by emergence of drug resistance, with ~40% of patients acquiring on-target ROS1 mutations at progression on crizotinib. NVL-520 is an investigational, brain-penetrant, and TRK-sparing ROS1 TKI designed to maintain activity for ROS1 mutations that confer resistance to approved therapies. To investigate suppression of resistance by NVL-520, we performed two N-ethyl-N-nitrosourea (ENU) mutagenesis screens and determined the frequency and identity of resistance mutations following treatment with NVL-520 versus earlier-generation ROS1 TKIs. Methods. Cells were treated with ENU for 1 day followed by crizotinib, entrectinib, repotrectinib, or NVL-520 for 28 days. Frequency of resistance indicates the fraction of wells that proliferated on day 28 despite continuous TKI treatment. Resistance mutations in the ROS1 kinase domain were identified by Sanger sequencing. Resistant clones harboring ROS1 kinase-domain mutations were expanded and used to profile the inhibitory activity of ROS1 TKIs. Results. The first ENU screen was performed on Ba/F3 CD74-ROS1, simulating a first-line treatment setting. NVL-520 at an unbound drug concentration (Cunbound) of 8.2 nM yielded 1 resistant clone among 480 replicate wells (frequency of resistance = 0.2%). The second screen was performed on Ba/F3 CD74-ROS1 G2032R, which is the most common point mutation after progression on previous-generation therapies, simulating a later-line treatment setting. NVL-520 at Cunbound of 41 nM yielded 6 resistant clones among 480 replicate wells (frequency of resistance = 1.3%). These NVL-520 concentrations (8.2 - 41 nM) were well within its average Cunbound of 166 nM at the recommended phase 2 dose (RP2D) of 100 mg once daily. None of the 7 resistant clones harbored ROS1 kinase-domain mutations, suggesting an off-target resistance mechanism. Crizotinib, entrectinib, and repotrectinib at their respective average human Cunbound yielded frequencies of resistance ranging from >33% to 100% in at least one of the screens, indicating that they were less effective than NVL-520 at suppressing resistance in these screens. Conclusion. NVL-520 was highly effective at suppressing resistance in ENU mutagenesis screens simulating first-line (CD74-ROS1) and later-line (CD74-ROS1 G2032R) treatment. The screens did not reveal any novel ROS1 mutations that could confer resistance to NVL-520 at the RP2D. The broad preclinical activity against ROS1 and diverse ROS1 mutations, together with its brain penetrance and TRK-sparing design, suggests NVL-520 as a potential best-in-class ROS1 therapy. Citation Format: Anupong Tangpeerachaikul, Franklin Gu, Henry E. Pelish. Mutagenesis screens support potential best-in-class profile for selective, brain-penetrant, and TRK-sparing ROS1 inhibitor NVL-520 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB182.